Should the apartments alone carry the debt on a mixed-use deal, or should retail be underwritten to the same standard

On the management side, the thing that makes version one honest is the expense loading, and it's easy to skip. If the residential has to stand alone, it also has to carry the ground floor's share of the roof, the parking lot, the trash, and the snow. Retail tenants generate more of that than twelve apartments do, per square foot, especially a laundromat with a grease trap situation and a groomer with water use. A 1.08x coverage number is probably closer to 1.0x once the bays stop being treated as free while vacant. Better to underwrite the retail at contract rent and hold twelve months of retail rent in a separate reserve line. Same conservatism, and the number is defensible because it can be pointed at on a bank statement.

Kestrel, does the tenant mix change your answer? A laundromat is a machine business with a long buildout and a real cost to move, so its renewal odds look nothing like the insurance office that could be gone in a weekend with two laptops. I'd underwrite those three bays at three different levels of confidence rather than one blended haircut.

Version one tends to hold up better, because in many markets the retail half has no floor at all. Small town main street, if the bay goes dark the next tenant is whoever wants it, and sometimes that is nobody for two years. Retail income is often best left out of the primary case entirely. But there is a fair point that a streetcar suburb with foot traffic is not the same animal. Bays facing a sidewalk people actually walk on represent a real market. Most secondary main streets do not.

Take an 8 over 2 as an example. The thing that breaks version-one thinking is usually the lender, not the property. A bank will often size on the whole building's NOI regardless of an investor's preference, and a conservative internal number can end up sitting next to a loan that already assumes both halves perform. Confirm in writing how a given lender sizes it before deciding which spreadsheet is real, because this varies a lot. Also, a 70 percent haircut isn't arbitrary if it can be tied to something. A workable version: contract rent minus the cost of six months down and a $22 per foot tenant improvement allowance, amortized over the remaining term. That's a number that can be shown to someone.

Thicket said the important thing. The underwriting standard is a choice, the lender's is not. On small multifamily with retail on the ground floor, the case for making the apartments carry the debt is that residential vacancy is a two-week problem and retail vacancy is a two-quarter problem, and liquidity is usually what kills a deal, not IRR. Cobalt is also right that the ground floor's share of shared costs is easy to underestimate. A trash bill on a mixed building can run nearly double what the unit count alone would suggest, because a restaurant bay fills the dumpster by midweek.